Environmental Effects on Cu/SiO 2 and Cu/Ti/SiO 2 Thin Film Adhesion
- PDF / 1,735,517 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 47 Downloads / 234 Views
=4J/m2 [5]was higher than typical metal/metal interfacial surface energies but
comparable to these for A120 3 on W, Mo and Nb as obtained by contact angle measurements [9]. One possibility is that Tit 2 was formed prior to Cu deposition [5] and lo=4J/m2 corresponds to Cu/TiO 2 rather than a Cu/Ti interface. Higher surface roughness of Ti or Tit
F0 [51.
2 as
compared to that for SiO 2can also be a reason for an observed increase in
269 Mat. Res. Soc. Symp. Proc. Vol. 563 ©1999 Materials Research Society
Table L. Techniques utilized for film thickness and grain size determination as well as residual stress measurements and yield stress evaluation are detailed elsewhere [5]. Film thickness (nm)
Grain Size (am)
200
110
0,, (GPa)
Table 1. Cu film characteristics [4]
1.37
500 130 0.89 1000 150 0.84 Hydrogen charging involved a IM NaOH solution with a Pt electrode as an anode. Films ranging from 80 to 3000 nm have been evaluated qualitatively by surface monitoring during hydrogen charging. Additionally, AFM observations were carried out following charging. Attention was focused on the differences in hydrogen induced blister formation for Cu and Cu/Ti systems. For a quantitative adhesion assessment, 200, 500 and 1000 nm thick Cu and Cu/Ti films have been examined under indentation loading. First, interfacial toughness was determined for non-charged samples. Next, films were tested immediately after -80 sec of hydrogen charging at 60 mA/cm 2 . Nanoindentation tests have been carried out with the 900 conical indenter of approximately 1 gtm tip radius. Indentation loads of 50, 100, 200 300 and 400 mN were applied with the MMT, a device analogous to an IBM micromechanical tester described elsewhere [10]. Adhesion calculations were based on the Marshall and Evans analysis[ 1I]. Here, critical strain energy release rate, G•,, is defined as a function of the following parameters: E, Young's modulus; v, Poisson's ratio; OR, residual stress in a film; R, delarnination radius and V0 , volume displaced by indentation. With the indentation depth exceeding film thickness, V0 can be 2R
determined as a volume of a truncated cone with the height equal to the film thickness and a base "equalto an indentation contact radius as shown in Fig. 1. Due to elastic recovery, an angle P is not equal to the indenter included angle. However, for a large contact radius, elastic recovery may be neglected. Delamination and
Sam 0
2b
contact radii as well as angles P3 have been
Figure 1. Schematic of an indentation induced delamination,
evaluated from AFM scans of indentation induced blisters as shown in Fig. 2. Effects of radial cracking and indentation induced pile-up are discussed in the APPENDIX. Note that analysis does not require indentation curves.
S_bst
so
n3.3
02
10-2
2a
2R
-----0.0 0 20 40 ptm b) ,040 5o, so 10 zo Figure 2. AFM measurements of delamination parameters. a) Blister image. b) Height profile
a)
_ 0
270
RESULTS AND DISCUSSION Blisters Induced During Hydrogen Charging A summary of Cu surface observations
Data Loading...
